X-UIDL: 817495306.000 Date: Mon, 27 Nov 1995 17:46:34 +0000 (GMT) From: Neil Geddes To: slacbuk@rl.ac.uk Cc: bob_jacobsen@lbl.gov, shi@llnl.gov, marsiske@slac.stanford.edu Subject: Last weeks "reconstruction" meeting in the UK "Minutes" of first UK reconstruction meeting - November 23, RAL Present: Neil Geddes, Paul Dauncey, Tim Adye, Gian Gopal, Steve Playfer, Adil Hasan, Damian Johnson, John Allison, Bob Jacobsen (Phone) Apologies from: Steve McMahon, Yuanning Gao, Mike Green, Neil Dyce, Paul Harrison These are intended to be a very brief account of what took place. The are in no way "official" and have not been approved by anyone else present. The meeting started with the proposal and adoption of the following agenda 1/ Brief introduction and overview of (reconstruction) computing by Neil 2/ Phone presentation by Bob Jacobsen on organisation of and plans for BaBar reconstruction 3/ Discussion of where to go next. 1/ Introduction (Neil Geddes) Main aim of meeting is to fill people in on where things currently are and to come up with some ideas of things that can be realistically achieved before the Paris meeting. Reconstruction efforts are now organised into four areas; Tracking, neutrals, particle id and common code. There has been a lot of effort most recently on the firts and last of these two; primarily as most other things depend upon them. There has been very little reconstruction effort to date (within the OO/C++ reconstruction framework) on the neutral reconstruction. Previous work, within bbsim, includes work by Al Eisner and Georges Vasseur on calorimeter clustering and current work includes implementation of digitisation code into the framework by Xiaorong Shi. It is intended to devote significant time at the reconstruction workshop in Paris to making progress in neutral reconstruction. A specific goal here is the specification of the interfaces for things like "calorimeter clusters", "neutral clusters" (are these the same ?) etc. 2/ Reconstruction (Bob Jacobsen) Bob Joined us by phone for about an hour. His slides can be found on the BaBar-UK home page ( http://hepunx.rl.ac.uk/BaBar/ ). There were several questions during and after Bobs presentation, but sadly I did not take any notes. Please read his slides. 3/ Discussion Steve Playfer had already spent some time before the meeting thinking about the design of the calormeter/neutral reconstruction software, based on his CLEO experience. He presented soime initial ideas towards defining the bahaviou and properties of CsI "clusters. (the following is intended only to give a flavour of the presentation/discussion - please contact Steve P. for more info) Basic idea of calorimeter clusters is straightforward - contiguous regions of block with deposited energy. What is "contiguous ? Deposited energy => energy above some threshold. Also have to handle split-offs and dead blocks. Data typically required by (defining) a cluster: Energy Centre of gravity Invaiant mass ratios of block energies (eg. E1/E9, E9/E25) Shower symmetry (size in theta and phi) secondary maxima some of these could be calculated as and when required. FOr how long must we keep a list of contributing blocks ? A track matching module would act on these clusters to prduce three new lists: charged tracks neutral clusters unmatched tracks (ie tracks with no cluster associated) "final" analysese will actually want to deal with "physicsl objects" (electrons, photons, pi0, eta, K0). Should the neutral reconstruction aim at producing lists of these objects (overlapping or non overlapping lists). An alternaitve would be a list of neutral objects which can answer questions such as "what is you energy under the electron hypothesis" or "what is the probability of you being a 280MeV muon". Steve plans to continue this work and to hopefully come up with the first steps towards interface definition for clusters classes over the coming weeks. Others present volunteered the following.... Paul Dauncey to look at the digitisation definitions being implemeneted by Xiaorong SHi to check that they look like what we might expect from the real electronics and to modify this as required to produce the trigger digitisations, which will also be required. Adil Hasan, needs to learn C++ and will attempt to do this by initially implementing Georges Vasseurs clustering algorithms as a Framework Module. Adil will communicate with Stave about anything he learns about the cluster class interface while doing this. It was generally agreed that we should attempt to meet by phone with interested parties elsewhere in the collaboration, sooner rather than later. (in particular Xiaorong Shi and Helmust Marsiske). Bob J. would also like to attend, which made 3pm on Tuesday November 28 the only practical time this week. This is not possible for Helmut. So we will attempt to put together a meeting for some time later this week of next week. Neil Geddes +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Neil Geddes | Tel: ++ 44 1235 445396 1-92, Building R1 | Rutherford Appleton lab. | Fax: ++ 44 1235 446733 Chilton, Didcot | Oxon OX11 0XQ | email: N.I.Geddes@rutherford.ac.uk England | +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++